A low-cost, fully electric, fully self-driving RV the size of a shipping container

1. The premise

A standard ISO container is a brutal design brief: about 2.4 m wide, roughly 2.6 m high externally for a high-cube, and either 6 m (20′) or 12 m (40′) long. As a living volume that is cramped; as a road vehicle envelope it is already wider and taller than many passenger cars, and a full 40′ box is truck territory.

“Fully electric” means no diesel generator as a crutch for range. “Fully self-driving” means the hard version of autonomy — not adaptive cruise on a good day, but the expectation that the vehicle can move people and their stuff without a licensed pilot in the loop for ordinary trips. “Low cost” is the third constraint that makes the first two interesting: if the only way to close the triangle is a six-figure battery and a robotaxi-grade sensor suite, you have a lab demo, not a recreational product.

2. Form factor: container as constraint

Road rules, not container standards, decide what you can actually drive. In much of North America, legal width without special permits is about 2.6 m (8′6″); a container’s ~2.44 m exterior width fits, but once you add mirrors, fenders, insulation, and structure, the “shipping box on wheels” fantasy meets chassis and body engineering immediately.

Height is similar: high-cube exterior height is already near the comfort zone for bridges and parking structures. Length is a spectrum: a 20′ class living module is more “tiny home on a purpose-built EV skateboard”; a 40′ class vehicle is closer to a Class A coach or box truck in mass, tire count, and licensing.

Mass is the quiet killer. Steel containers are heavy before you add battery, motors, suspension, furniture, water, and people. A low-cost design almost certainly does not reuse a steel intermodal box as the primary structure; it uses the dimensions as a size target while building a lighter monocoque or spaceframe body meant for the road.

• Envelope: think high-cube 20′ first; 40′ only if you accept truck-like complexity.
• Body: composite or aluminum skin over a purpose-built EV chassis, not a raw container.
• Interior: fixed wet cell + transformable living space; volume is the scarce resource.

3. Energy: batteries vs a rolling house

A boxy RV is aerodynamically rude. At highway speed, drag dominates. Ballpark consumption for a large, tall electric vehicle is often quoted in the 300–600 Wh/km range depending on speed, mass, and weather — several times a compact car. That means range is expensive in both dollars and kilograms.

Order-of-magnitude pack math (pack-level, not cell fantasy): roughly 150–180 Wh/kg usable pack energy density and on the order of $100–150 / kWh pack cost in the mid-2020s, trending down but not free. A 100 kWh pack is then hundreds of kilograms and five figures of cost before integration, thermal management, and structure.

Use the estimator below (implemented in Rust WASM) to feel the trade: larger packs buy range and cost mass and money; they do not fix the aero tax of a house-shaped vehicle.

Charging and duty cycle matter as much as pack size. A recreational vehicle that mostly sits at a campground on shore power has a different battery story than one expected to hop 400 km between national parks every day. “Low cost” might mean modest range + ubiquitous destination charging rather than Tesla-like road-trip buffer for a brick.

4. Autonomy: the expensive last 1%

Highway lane-keeping and adaptive cruise are commodity-ish. Full self-driving in the sense of unsupervised operation on public roads — weather, construction, pedestrians, parking lots, dirt access roads to campsites — is still the frontier where companies burn billions.

Sensor suites (cameras, radar, optional lidar), compute, mapping, validation, and liability dominate cost long after the motors are chosen. For an RV, low-speed yard maneuvering and remote tele-assist might be more valuable than unsupervised city driving, and far cheaper to make honest. Regulatory regimes also differ by jurisdiction; a product that is legal to sell as “fully self-driving” in one market may be “driver assist only” in another.

• Cheap-ish: ADAS for highway fatigue reduction and parking assist.
• Hard: unsupervised door-to-door trips with sleeping occupants.
• Maybe FOSS-adjacent: open tooling for mapping, simulation, and driver-out-of-loop monitoring — not the full stack on day one.

5. Cost stack: where “low cost” breaks

A crude stack for a container-scale EV RV might look like:

1. Rolling chassis — motors, inverters, suspension, brakes, thermal.
2. Battery pack — cells, modules, BMS, enclosure, crash structure.
3. Body & interior — insulated shell, glazing, wet cell, furniture.
4. Autonomy & electronics — compute, sensors, wiring, HMI.
5. Certification & support — crash, emissions-adjacent rules, service network.

Battery and autonomy are the two cliffs. Body and chassis can be value-engineered; liability and validation cannot be wishcast away. “Low cost” almost certainly means narrowing the mission: shorter range, human still responsible for driving for years, shared platforms with commercial EV vans or skateboards, and interior kit that is modular rather than yacht-grade.

6. Open questions and a FOSS angle

What could actually be open?

• Software — energy management UIs, trip planners tuned for RV duty cycles, open diagnostics, non-cloud lock-in for vehicle telemetry the owner actually owns.
• Design — published dimensions, mass budgets, and interior modules others can fork (the way open hardware frames spread in other domains).
• Not magically open — cell factories, full autonomy stacks trained on proprietary fleets, and type approval. Fossall’s bet is not that the whole vehicle is a git repo; it is that the closed middle of mobility should shrink.

A shipping-container-sized electric, self-driving RV that is cheap is probably impossible if you demand all three adjectives at once in their strongest form. A container-scale electric camper with strong driver assistance, honest range, and open software around it is a much more interesting near-term target — and a better place to start building in public.